?: Heater supply PSU for C12 clone

Hi,
I have been looking at the schematic for the Matador/Chunger C12 clone as an inspiration for a DIY project I have been working on:

http://musicalsparks.com/images/c12/c12_schematics_v3.pdf

and I am wondering what the reasoning is behind the choice of a 4700uF value for C6, C7, and C8 in the heater supply.

I am not trying to be difficult. I don't have very much experience with regulated supplies so this is a sincere question. The C12 clone design seems like a hybrid of old fashioned unregulated layouts using lots of RC filtering with a regulator circuit placed right in the middle.

I have been looking at the data sheet for Fairchild's LM317 and it seems as if a 0.1uF filter cap is sufficient.

I am looking at the price difference between 4700uF caps and other choices, 1000uF for example, the 1000 is half the price of the 4700 and both seem like serious over kill.

I am also attracted to the idea of using smaller components such as are available for a 1000uF spec.

Finally I am wondering why there are 3 filter stages in front of the LM317 when it doesn't seem to matter.

Please, correct me if I am wrong, is there a benefit to including all the extra stages of high value RC filtering?

Thank You for any help you can offer.


 

0.1uF is recommended close to the regulator if the bulk capacitance is some distance away (if memory serves, it's 2-3inches).  It is *not* sufficient by itself for a heater supply though.

So right after the rectification diodes is a lot of ripple:  an LM317 can never output a voltage higher than it's input (minus the dropout):  if the current requirements are high, the voltage pre-regulator can drop rapidly, and the regulator will follow it downward and all bets are off.  Bulk capacitance pre-regulation helps greatly with this:  it can serve current to the load during the lulls in the incoming AC waveform and keep the regulator input voltage happily above it's dropout point.

Now having said all that, in this design, there is generally 7V across the regulator at all times.  Since it's a heater supply, load current is mostly constant as well.  A good rule-of-thumb is 1000uF for every 1 amp of load current.  A C12 is basically a constant 350mA load, so even a 470uF would probably be ok in most cases.

I added the CRCRC filter stages to experiment with passive RC filtered heater supplies, but never got around to it.  I choose 4700uF because it's what Mouser had in large supply at the time.

If you really want to strip it down to bare bones:  drop C6 and C10 down to 1,000uF, jumper across R7 and R8, and remove C7.  It will probably work just fine.

Hi Matador,
Thanks very much for such a thorough and easy to comprehend explanation.

The idea about the ripple pushing the limits of the in/out difference was not something I would have imagined, and you explanation is very helpful.

Thank You!